Research on the Path of Enhancing Intercultural Communication Competence of Students in Applied Colleges and Universities Supported by Intelligent Learning System

Cultivating the spirit of cultural inheritance and cultural innovation among college students is one of the historical missions of contemporary higher education. However, for a long time, the cultivation of foreign language majors in China has mostly emphasized on the training of language skills, and the teaching of culture only focuses on how to understand and absorb foreign cultures in order to communicate with the target language group, ignoring the decisive role of foreign language learning in the international dissemination of local cultures, and not paying attention to cultural communication ability as a part of foreign language ability. The role of foreign language learning in the international dissemination of local culture has been neglected, and cultural communication ability has not been emphasized as a part of foreign language ability. In the context of globalization, it is the requirement of the times to improve the cultural literacy, cultural communication awareness and ability of foreign language majors [1-2]. In recent years, with the concept of “cultural power” put forward by the state, the foreign communication of Chinese culture and the foreign language ability of the country have become the focus of attention of many scholars, and the cultivation of the cultural communication ability of foreign language majors, which constitutes the main force of the foreign language ability of the country, should also be given enough attention [3].

With the rapid development of information technology, the application of virtual simulation and artificial intelligence technology in the field of foreign language teaching is rapidly becoming a research hot spot. By creating an immersive learning environment, virtual simulation technology enables learners to practice and experience language learning in simulated scenarios, thus significantly improving their language learning efficiency and interest. The main task of foreign language teaching is to cultivate students’ intercultural communicative competence, i.e., the comprehensive competence consisting of verbal communication, nonverbal communication, etc. It is advocated that classroom teaching be placed in an intercultural communicative environment [4-5]. The ultimate goal of cross-cultural foreign language teaching is to cultivate students’ cross-cultural competence, so that they can have an international perspective and Chinese sentiment, and become cross-cultural talents who can freely cross cultural boundaries. Therefore, in virtual simulation foreign language teaching, intercultural communication elements can be presented from more modalities with far-reaching significance [6].

With the popularization of globalization and all-round communication, the status of intercultural communicative competence has become more and more prominent, so there are many studies and analyses on intercultural communicative competence, which involve the influencing factors of intercultural communicative competence, the cultivation path of intercultural communicative competence, etc. Jiang, S et al. elaborated on the introduction of intercultural background learning in the process of intercultural communicative language competence cultivation and analyzed the practical effects of multimodal change and the practical effect of culture introduction teaching method, which makes a positive contribution to the enhancement of students’ intercultural communicative competence [7]. Malazonia, D et al. examined how different cultural factors affect the development of students’ intercultural competence, pointed out that students’ negative attitudes towards intercultural communication were attributed to the insufficient intercultural communication climate and related guidance in schools, and argued that schools need to take responsibility for the development of students’ intercultural communication competence [8]. Wang, Y et al. combined semi-structured interviews as well as questionnaire methods to reveal the differences in oral communication barriers between Chinese and Iranian English majors and conducted a related analysis to clarify the importance of emotions for intercultural oral communication [9]. Lantz-Deaton, C et al. empirical analysis discusses that students’ experience of engaging in intercultural communication is not enough to enhance students’ intercultural communication competence, pointing out that it is the school’s policy of teaching intercultural competence and targeted practice that can effectively develop students’ intercultural communication competence [10]. Strohmeier, D et al. used a self-regulated school equation model to analyze the development of an experimental intercultural teaching class, and the results of the study showed that intercultural learning goals, intercultural self-efficacy for learning, and intrinsic interest in intercultural learning are prerequisites for the development of intercultural communicative competence, and explored in detail the practical effects and value of the theory of intercultural self-assessment [11].

Artificial intelligence technology in education has always been a popular research in the field of education, and many experts and scholars have carried out a lot of research on the design of teaching methods, value potential and potential threats of artificial intelligence technology in education. Popenici, S. A et al. explored the cutting-edge practices and related research on AI technologies in higher education and analyzed the current obstacles and challenges associated with the introduction of AI technologies in higher education, and provided an outlook on potential future applications and research directions [12]. Alam, A et al. based on the analysis of educational practice cases of AI technology, learned that the application areas of AI technology include teaching assessment, intelligent robot tutoring, smart campus and virtual classroom, etc., and deeply analyzed the potential of AI technology in promoting teaching reform [13]. Qadir, J et al. revealed that ChatGPT, an AI technology, brings personalized learning experience for student learning along with ethical issues such as immorality and dishonesty, arguing that there is a need for a deeper understanding of the impacts brought by AI technology, and continuous optimization and improvement of the practice of AI technology in education [14]. Rudolph, J et al. analyzed the positioning and practical direction of chatbots based on AI technology in higher education, and made a dialectical analysis of the opportunities and threats posed by AI technology, which made a positive contribution to the construction of information intelligence in higher education [15].

The study proposes relevant factors that affect the intercultural communication competence of college students using the Intelligent Language Support System (ILSS), combined with the TAM model and UTAUT model. Subsequently, students enrolled in an applied university in Guangzhou City were selected as the research subjects, and a questionnaire was used to collect data, and structural equations were used to process the collected data. Finally, the levels of linguistic competence, cultural competence, communicative competence, and professional competence of the sample, as well as the differences in different genders and academic qualifications, were studied. On this basis, structural equations were utilized to test the proposed hypotheses and explain the mechanism of improving intercultural communication competence of students in applied universities supported by the intelligent learning system.

2

Study design

2.1

Intelligent Language Learning System

2.1.1

Crawler-based Data Processing Methods for Language Learning

Language learning data refers to digital information that is used to help learn a language, and this data usually includes vocabulary, grammar, pronunciation, and phrases. Crawling techniques can be used to automatically obtain information from language learning data, such as automatically downloading and parsing vocabulary and grammar rules from language learning websites. The study uses a hierarchical traversal algorithm to traverse a directed graph to obtain learning data from web pages. In a directed graph, each node represents a web resource, while each edge represents a link between two nodes. By traversing the nodes and edges, the algorithm can determine the type of each node (e.g., text, image, or video) and the linking relationship between them. The calculation of the degree of association between a collection of two entities on a web page is shown in equation (1).

(1)

C o r r e l (A, B) = \frac{| {C : A \cup B, C \in D} |}{| D |}

In equation (1), the two entity sets are A and B, and the correlation between them is correl(A,B). The web page entity exclusive set is D, and its internal item set is C. The credibility between the two entity sets is calculated as shown in equation (2).

(2)

C o n f i d (A, B) = \frac{| {C : A \cup B, C \in D} |}{| C : A \subseteq C, C \in D |}

In Eq. (2), the confidence level between entity sets A and B is confid(A,B). The item set is a frequent item set, which needs to satisfy two conditions: the first condition is that the relevance value of the item set is greater than or equal to the minimum threshold value, and the second condition is that the relevance value of the item set is greater than or equal to the value of the item relevance support.

2.1.2

Intelligent Speech Q&A Algorithm Design

Intelligent Speech Q&A algorithm is an intelligent algorithm that utilizes natural language processing and machine learning techniques to process speech input and answer user questions. The algorithm usually converts speech input into text and pre-processes the text with word division, lexical annotation, word sense disambiguation, and so on. Then, the system will extract the key information according to the user’s question, query the database for the answer, and finally generate an appropriate answer and output it to the user.

For the user’s voice input, the intelligent voice Q&A algorithm uses the similarity calculation model to perform word segmentation based on template matching. The frequency of each word occurrence needs to be counted and the inverse document word frequency of the word is calculated. Then, the samples are classified and divided by attribute features using Bayesian algorithms to generate different Bayesian models. Next, the distance between two samples is determined by similarity calculations to understand the similarities or differences between them. In the similarity calculation model, Euclidean distance, Manhattan distance, Minkowski distance, and cosine similarity can be used. Finally, according to the cosine similarity algorithm and word frequency-inverse document word frequency calculation results, and based on the maximum value of cosine, identify the corresponding entity relationship to generate the answer. The word frequency calculation formula is shown in equation (3).

(3)

T F = \frac{N_{t}}{T_{w}}

In equation (3), the word frequency is TF, the number of times a word appears in the text is N_t, and the total number of words in the text is T_w. The formula for calculating the word frequency of the inverse document is shown in equation (4).

(4)

I D F = \log (\frac{T_{d}}{N_{t d}})

In Eq. (4), the inverse document word frequency is IDF, the total number of documents in the corpus is T_d, and the number of documents containing a certain word is N_td. The word frequency-inverse document word frequency formula is shown in Eq. (5).

(5)

T F - I D F = T F \times I D F

In Eq. (5), the word frequency-inverse document word frequency is TF – IDF and its value is the product of the word frequency and the inverse document word frequency. In the TF – IDF-model, each document is represented as a vector that contains the frequency of all keywords in the document or the number of times the keyword appears in the text. These vectors are usually represented as a two-dimensional matrix, where the row vectors represent all the keywords in the document and the column vectors represent the frequency of each keyword or the number of times the keyword appears in the text. Each keyword has a corresponding weight which indicates the importance of that keyword in the document. The cosine similarity is calculated as shown in equation (6).

(6)

S i m (X, Y) = \cos (θ) = \frac{X \cdot Y}{‖ X ‖ ‖ Y ‖}

In Eq. (6), the two sample vectors are X and Y, the cosine similarity between them is sim(X,Y), and the angle between them is θ. Cosine similarity is a metric used to measure the degree of similarity between two vectors. It determines the similarity of two vectors by calculating the cosine of the angle between them. In a vector space model, each vector represents a feature vector of a document or a word, where each dimension represents a feature or attribute. Cosine similarity measures the similarity of two vectors by calculating the angle between them, the smaller the angle, the larger the cosine similarity, indicating that the two vectors are more similar [16].

2.1.3

Intelligent language learning system design

Intelligent language learning system is a system that utilizes artificial intelligence and natural language processing technology to help users learn languages. It can interact with users through speech recognition, speech synthesis, text analysis, and other technologies to provide personalized learning content and suggestions. Based on this, the research integrates crawler technology and an intelligent voice Q&A algorithm to build an intelligent language learning system. The structure of the intelligent language learning system is shown in Figure 1.

Intelligent language learning system is a software system that can help users learn and improve their language skills. The system uses crawler technology, entity relations, data storage, and voice answering to provide an interesting and personalized way of learning. In addition, the system can store the user’s learning materials and resources so that they can easily be accessed and reviewed. The system allows users to perform voice exercises to practice pronunciation, grammar, and communication skills. The system can also provide a personalized learning experience by answering the user’s questions using voice response technology.

2.2

Research Models and Assumptions

2.2.1

Research models

1)

Technology Acceptance Model (TAM)

The TAM model suggests that there are two key factors that influence people’s willingness to use computers, namely perceived usefulness and perceived ease of use. Among them, perceived usefulness is mainly used to reflect the extent to which users believe that using a mobile device improves their learning or work performance. Perceived ease of use, on the other hand, refers to the degree to which users find it easy to use mobile devices [17].

2)

Technology Acceptance and Use Integration Model (UTAUT)

Taking the ATM model as the theoretical basis and synthesizing more than 20 variables from eight models, such as the Technology Task Fit Model (referred to as the TTF model) and the Innovation Diffusion Theory (referred to as the IDT model), the Integration of Technology Acceptance and Use Model (referred to as the UTAUT model) was finally proposed after measurement and testing. The test results also show that the explanatory power of the UTAUT model for technology acceptance is close to 70%, and is widely regarded, as one of the powerful theoretical tools for predicting and explaining the acceptance of information technology by an individual or an organization,.

2.2.2

Research hypotheses

Referring to the findings of previous studies and inspired by the TAM model and the UTAUT model, this study proposes the following hypotheses:

H1: Perceived usefulness has a positive effect on students’ intercultural communication competence after accepting an intelligent learning system.

H2: Perceived validity has a positive effect on students’ intercultural communication competence after receiving an intelligent learning system.

H3: There is a positive effect of community influence on students’ acceptance of intercultural communication competence after an intelligent learning system.

H4: There is a positive effect of enabling conditions on students’ intercultural communication competence after an intelligent learning system.

H5: Perceived enjoyment has a positive effect on students’ intercultural communication competence after an intelligent learning system.

H6: There is a positive effect of learning satisfaction on students’ intercultural communication competence after receiving the intelligent learning system.

H7: There is a positive effect of the intelligent learning system on students’ intercultural communication competence after receiving the intelligent learning system.

H8: The intelligent learning system has a positive effect on perceived usefulness.

H9: The intelligent learning system has a positive effect on perceived effectiveness.

H10: The intelligent learning system has a positive effect on perceived ease of use.

H11: There is a positive effect of intelligent learning system on learning satisfaction.

2.3

Research methodology

2.3.1

Structural equation modeling analysis

Structural equation modeling (SEM) is a multivariate statistical method that combines factor analysis and path analysis to verify and analyze the complex relationships between variables through mathematical models, and its significant advantage is that it can deal with the complex path relationships that exist in multiple variables [18].

Structural equation modeling differs from traditional statistical methods mainly in the principle of operation, where the operating pointer of the latter is a specific variable, whereas the operating pointer of structural equation modeling is the covariance matrix that describes the relationship between the variables. The principle is to test the similarity between the actual covariance Σ and the sample covariance Σ(θ), i.e., to try to achieve the value of the fitted function |Σ – Σ(θ)| tends to be close to zero, therefore, the structural equation modeling is also called covariance modeling.

1)

Variable analysis

Variables are the basis of statistical analysis, and structural equation modeling has its own set of unique variable system, as follows: (1)

Latent variables:

Latent variables, also called hidden variables, concepts or factors, refer to specific variables in the sample that cannot be obtained through direct observation (e.g., ξ₁, η₁, η₂ in Fig. 2), and can be divided into two kinds of latent variables according to the difference in the influence relationship between the variables: one is exogenous latent variables, which refers to the variables that only have an effect on other variables in the model and are not affected by the other variables, i.e., the variables that only play an explanatory role in the model, which is usually expressed as ξ (e.g., ξ₁ in Fig. 2). The other is endogenous latent variable, which refers to the latent variable that is affected by any latent variable in the model, and is usually denoted by η (e.g., η₁ and η₂ in Fig. 2). It is worth noting that those that are affected by other latent variables as well as those that act on other latent variables, like Fig. 2 η₁, are also endogenous latent variables, which can be called mediator variables, and mediate the role between the two latent variables.

(2)

The residual term

The residual term refers to the error that cannot be explained in the model or exists in the variables themselves. There are two main types of errors in structural equation modeling: one is the error that exists in the system itself, i.e., the error caused by the endogenous latent variables that cannot be partially explained by the exogenous variables, which is called the systematic error and is usually denoted by ζ (e.g., ζ₁…ζ₃ in Fig. 2), and the second is the error that occurs during the measurement process, i.e., the error that occurs in the measurement of the observed variables, which is called the stochastic error, which is denoted by ε and δ, respectively, and is denoted by y and x of the observation errors (e.g., ε₁…ε₃ and δ₁…δ₅ in Fig. 2).

2)

Path analysis

Path analysis is an important index for structural equation modeling to analyze the impact relationship, including path diagram, path coefficient, and impact analysis of three parts, as follows: (1)

Path diagram: the path diagram is the legend that visualizes the path relationship between variables, and the influence relationship between variables is called path relationship in the model, according to the different kinds of variables, the path relationship can be divided into two kinds: latent variable → latent variable, latent variable → observed variable.

(2)

Path coefficient: the value on the path relationship is called the path coefficient, which is used to reflect the intensity of influence between the variables, and the standardized path coefficient should take the value of (0,1), and the closer the value is to 1, which means that the intensity of influence between the variables is greater. According to the different path relationships, there are two kinds of path coefficients: one is the path coefficient between latent variables to reflect the intensity of the influence between latent variables (such as Figure 2, h₁, h₂, h₃); the second is the path coefficient between latent variables and observed variables, which is called the factor loading coefficient, and it is used to reflect the intensity of influence of the observed variables on the corresponding latent variables (such as Figure 2, λ_x₁₀, λ_y₁, etc.).

3)

Measurement model

Measurement model refers to the model that demonstrates the relationship between each latent variable and the observed variables, and its principle of action belongs to the process of validation factor analysis, indirectly portraying the implicit latent variables through the introduction of the explicit observed variables, and the structural equation model contains two or more measurement models according to the number of latent variables. The mathematical form of measurement models is usually matrix equations, which are called measurement equations, as in equation (7): (7) ${\begin{array}{l} x = Λ_{x} ξ + δ \\ y = Λ η η + ε \end{array}$

In Eq. (7), $\begin{matrix} x = (\begin{array}{l} x_{1} \\ x_{2} \\ x_{3} \end{array}), & ξ = (ξ_{1}), & Λ_{x} = (\begin{array}{l} λ_{x_{11}} \\ λ_{x_{22}} \\ λ_{x_{1}} \end{array}), & δ = (\begin{array}{l} δ_{1} \\ δ_{2} \\ δ_{3} \end{array}) . & Λ_{x} \end{matrix}$ denote the coefficient matrices between the exogenous latent variable ξ and the observed variables x₁, x₂, and x₃. Similarly, $\begin{matrix} y = (\begin{matrix} y_{1} \\ ⋮ \\ y_{5} \end{matrix}), & η = (\begin{matrix} η_{1} \\ η_{2} \end{matrix}), & Λ y = (\begin{matrix} λ_{y_{11}} & 0 \\ λ_{y_{2}} & 0 \\ λ_{y_{13}} & 0 \\ 0 & λ_{y_{24}} \\ 0 & λ_{y_{5}} \end{matrix}) ε = (\begin{matrix} ε_{1} \\ ⋮ \\ ε_{5} \end{matrix}) . & Λ_{y} \end{matrix}$ denote the coefficient matrix between the endogenous latent variable η and the observed variable y₁…y₅, and Λ_x and Λ_x are referred to as the factor loading coefficient matrix.

4)

Structural model

Structural model refers to the model that demonstrates the relationship between the latent variables and interprets the influence relationship between the latent variables through path analysis, and each structural equation model contains only one structural model, and the mathematical form of the structural model is also a matrix equation called structural equation, as in equation (8): (8) $η = B η + Γ ξ + ζ$

In equation (8), $\begin{matrix} η = (\begin{matrix} η_{1} \\ η_{2} \end{matrix}), & ξ = (ξ_{1}), & ζ = (\begin{matrix} ζ_{1} \\ ζ_{2} \end{matrix}), & B = (\begin{matrix} 0 & h_{2} \\ 0 & 0 \end{matrix}), & Γ = (\begin{array}{l} h_{1} & h_{3} \end{array}) . & B \end{matrix}$ denotes the coefficient matrix between endogenous latent variables η, and Γ denotes the coefficient matrix between exogenous latent variables ξ and multiple endogenous latent variables η.

Structural equation modeling is a complex model consisting of a structural model together with multiple measurement models, so the principle equation of structural equation modeling, as in Eq. (9): (9) ${\begin{cases} x = Λ_{x} ξ + δ \\ y = Λ, η + ε \\ η = B η + Γ ξ + ζ \end{cases}$

2.3.2

Structural equation modeling analysis process

According to the principle of the role of structural equation modeling, the analytical process of the model includes the following three stages: model development, model setting and assessment, and model determination. The first is the model development stage, which sorts out the theoretical basis of the study, sets the framework of the content to be studied according to the theory, and provides a solid foundation for the model assumptions. Second is the model setting and assessment stage, which determines the latent variables and the relationship between the latent variables in the study based on the theoretical assumptions, and designs the theoretical model by introducing the form of observational variables to portray the latent variables. The parameters to be assessed are identified based on the set relationships, and then the model is identified and assessed, and the reasonableness of the model reliability and validity is tested. Finally, the model determination stage, the sample data will be substituted into the theoretical model to fit the model, and the fitted model will be used for parameter estimation, and the final structural equation model can be determined after the estimation results are reasonable, on the contrary, the model needs to be revised and repeat the above steps until it is reasonable. The process of analyzing the structural equation model is shown in Figure 3.

2.4

Data sources and processing

The survey respondents selected for this study were from students enrolled in an applied university in Guangzhou City, and the method of questionnaire survey was adopted. 560 questionnaires were randomly distributed and 530 valid questionnaires were recovered, with a valid recovery rate of 94.6%. Among them, 480 students had cross-cultural communication experience, accounting for 90.6% of the valid questionnaires. The questionnaire used SPSS25.0 to examine the students’ linguistic competence, cultural competence, communicative competence, and professional competence.

3

Empirical analysis

3.1

Results of the survey on the current status of students’ intercultural communication competence

3.1.1

Sample descriptive statistics

The descriptive statistics of 530 questionnaires were analyzed to find out the gender, subject category, education level, place of origin, and family location of the students. The results of the survey are shown in Table 1.

Table 1.

Basic descriptive statistics of individual characteristics

Individual characteristics	Categories	Frequency	Proportion
Gender	Male	318	60%
Gender	Female	212	40%
Subject category	Science class	173	32.6%
	Engineering class	78	14.7%
	Social sciences	171	32.3%
	Art class	108	20.4%
Educational background	Graduate student	355	67%
Educational background	Undergraduates	175	33%
Source area	Eastern region	212	40%
	Central region	187	35.3%
	Western region	131	24.7%
Come from	Town	279	52.6%
Come from	Countryside	251	47.4%

As shown in Table 1, among the 530 valid questionnaires surveyed, males totaled 318 questionnaires, accounting for 60%, and females totaled 212 questionnaires, accounting for 40%.

From the viewpoint of subject categories, the total number of science students is 173, accounting for 32.6%, the total number of engineering students is 78, accounting for 14.7%, the total number of social science students is 171, accounting for 32.3%, and the total number of humanities and arts students is 108, accounting for 20.4%.

In terms of educational level, students totaled 327, accounting for 76.2%. Undergraduate students totaled 102, accounting for 23.8%, indicating that most of the students who filled out the questionnaire were master’s degree students.

From the viewpoint of source region, 212 students came from the eastern region, accounting for 40%, 187 students came from the central region, accounting for 35.3%, and students from the western region totaled 131, accounting for 24.7%.

In terms of urban and rural sources, 279 students, or 52.6%, came from towns and cities, and 251 students, or 47.4%, came from rural areas.

3.1.2

Current status of students’ intercultural communication competence and its dimensions

The mean and standard deviation of students’ intercultural communication competence and its dimensions in a university in Guangzhou are shown in Table 2. According to the table, the overall mean score of students’ intercultural communication competence is 4.74, and the standard deviation is 0.58, which means that students’ intercultural communication competence is at the “average” level. Among the three dimensions of students’ intercultural communication competence, the mean value of the communicative competence dimension is 4.98 with a standard deviation of 0.72, the mean value of the linguistic competence dimension is 4.51 with a standard deviation of 0.76, and the mean value of the cultural competence dimension is 4.11 with a standard deviation of 0.81.

Table 2.

Descriptive statistics

Dimension	Mean	Standard deviation
Linguistic ability	4.51	0.76
Cultural ability	4.11	0.81
Communicative ability	4.98	0.72
Cross-cultural communication ability	4.74	0.58

3.1.3

Differential analysis of students’ intercultural communication competence

In this study, 429530 valid questionnaires were analyzed for differences using SPSS 25.0 statistical analysis software. The study utilized independent samples t-test to find out whether there is a significant difference between different gender, subject category, and educational qualification variables in students’ intercultural communication competence, its dimensions, and its factors. The one-way ANOVA method was utilized to test whether there is a significant difference between different gender and educational qualification variables in students’ intercultural communication competence, its dimensions, and its various factors.

1)

Gender differences

According to the difference of gender, the students are divided into two groups of male and female students, and the method of independent sample t-test is mainly used to analyze the significant differences between students of different genders in intercultural communication competence and its dimensions, and the specific results are shown in Table 3. In terms of the dimensions, the P-values of intercultural communication competence and the dimensions in the table are all greater than 0.05, indicating that there is no statistically significant difference between the students’ in the variable of gender.

Table 3.

Gender analysis

Variable	Gender	Case number	Mean value	Standard deviation	T	P
Linguistic ability	Male	318	3.454	0.822	-0.892	0.572
Linguistic ability	Female	212	3.508	0.755	-0.892	0.572
Cultural ability	Male	318	3.055	0.819	-2.131	0.437
Cultural ability	Female	212	3.123	0.706	-2.131	0.437
Communicative ability	Male	318	3.882	0.736	-2.641	0.234
Communicative ability	Female	212	3.974	0.616	-2.641	0.234
Cross-cultural communication ability	Male	318	3.463	0.67	-2.342	0.216
Cross-cultural communication ability	Female	212	3.533	0.563	-2.342	0.216

Note: ^* represents P<0.05, ^** represents P<0.01, and ^*** represents P<0.001.

2)

Differences in academic levels

The students investigated in this study were mainly divided into two categories: graduate students and undergraduate students, and the independent samples t-test was chosen to test whether there was a significant difference between the students’ intercultural communication competence and the dimensions at the academic level, and the results of the test are shown in Table 4.

Table 4.

The differences in the learning calendar

Dimension	Educational background	Case number	Mean value	Standard deviation	T	P
Linguistic ability	Undergraduates	355	3.423	0.713	-4.512	0.009^**
Linguistic ability	Graduate student	175	3.672	0.952	-4.512	0.009^**
Cultural ability	Undergraduates	355	3.052	0.85	-3.252	0.000^***
Cultural ability	Graduate student	175	3.217	0.261	-3.252	0.000^***
Communicative ability	Undergraduates	355	3.935	0.721	0.621	0.936
Communicative ability	Graduate student	175	3.918	0.506	0.621	0.936
Cross-cultural communication ability	Undergraduates	355	3.467	0.635	-4.352	0.051^*
Cross-cultural communication ability	Graduate student	175	3.607	0.539	-4.352	0.051^*

Note: indicates P<0.05,

indicates P<0.01, and

indicates P<0.001.

As can be seen from the table, in the dimension of linguistic competence, the mean value of undergraduate students is 3.423, and the mean value of graduate students is 3.672, with a T-value of -4.512, P=0.009, which passes the test of significance at the 0.01 level. In the dimension of cultural competence, the mean of undergraduate students was 3.052, and the mean of graduate students was 3.217, with a T-value of -3.252, P=0.000<0.001, which passed the test of significance at the level of 0.001. In the dimension of communicative competence, the T-value was 0.621, P=0.936, which did not pass the test of significance at the 0.05 level. In professional competence, the mean value of 3.467 for undergraduate research students and 3.607 for postgraduate students, with a T-value of -4.352, P=0.051<0.05, were significantly different at the 0.05 level. It indicates that there is a significant difference between students of different academic levels in the dimensions of intercultural communication competence, linguistic competence and cultural competence, and there is no significant difference in the dimension of communicative competence. Overall, the level of intercultural communication competence of graduate students is higher than the level of competence of undergraduate students.

3.2

Students’ intercultural communication competence pathway enhancement

3.2.1

Structural equation modeling

A total of nine research variables were set in this study to construct the model, which are: (1) The dependent variable is students’ cross-cultural communication competence (BI) after receiving the intelligent learning system. (2) Seven independent variables, including two core variables of perceived usefulness (PE) and perceived ease of use (PEU) in the TAM model theory, two core variables of community influence (SI) and facilitating conditions (FC) in the UTAUT model theory, and three core variables of perceived pleasantness (PJ), satisfaction with learning (SAT), and the intelligent learning system (CEE), which were set up based on the ternary interaction theory.

According to the calculated standard regression coefficient values range from 0.73 to 0.96, all the observed variables showed a strong positive correlation with the latent variables. The factor loading coefficients of the regression weighted results of the latent variables (independent variables) and the observed variables (dependent variables) were obtained to be in the range of 0.7 to 0.9, which indicated that the test of the observed variables as a measure of the latent variables was appropriate.

Distinguishing validity signifies the distinction between latent variables. There are two indicators of correlation: the average variance extracted AVE value (AVE value ≥ 0.5), and the square root of the average variance extracted AVE (AVE square root > correlation coefficient). The mean variance extracted AVE value was calculated from the factor loadings and was the sum of the squared factor loadings. The square root of the mean variance extracted AVE is greater than the absolute value of the correlation coefficient between the latent variables. The results of the discriminant validity test between latent variables are shown in Table 5.

Table 5.

The distinguishing validity table between the latent variables

	PE	PEU	FC	SI	CEE	PJ	BI
PE	0.675	--	--	--	--	--	--
PEU	0.84	0.668	--	--	--	--	--
FC	0.917	0.925	0.622	--	--	--	--
SI	0.924	0.965	0.963	0.588	--	--	--
CEE	0.863	0.917	0.913	0.966	0.725	--	--
PJ	0.92	0.959	0.951	0.974	0.96	0.62	--
BI	0.921	0.917	0.959	0.977	0.984	0.957	0.701
The square root of AVE	0.935	1.03	1.001	0.978	1.005	1.02	1.011

The results of the relevant indexes are shown in Table 6, from which it can be seen that GFI and AGFI are slightly lower than the critical value, and all the rest of the indexes meet the fitness standard, and the GFI value and AGFI value greater than 0.8 are up to standard. Therefore, all the standards are met this time.

Table 6.

Final equation model suitability test analysis results

Index type	Statistical inspection quantity	Ideal standard value	Test data	Model fitting judgment
Absolute fitting index	X₂	P<0.001	0.926	yes
	RMR value	<0.05	0.914	yes
	RMSEA value	<0.1(<.05 Goodness;<.08Good)	0.994	yes
	GFI value	>.90	0.993	yes
	AGFI value	>.90	0.994	yes
Relative fitting index	NFI value	>.90	0.923	yes
	RFI value	>.90	0.998	yes
	IFI value	>.90	0.857	yes
	TLI value	>.90	1.086	yes
	CFIvalue	>.90	0.926	yes
Contracted fit index	PGFI value	>.50	0.914	yes
	PNFI value	>.50	0.994	yes
	PCFI value	>.50	0.993	yes
	CMIN/DF	<5.00(<3.00 goodness)	1.045	yes

3.2.2

Research hypothesis testing

After obtaining the model with good fitness, the path coefficients of the 11 hypothesized paths were derived with the help of AMOS 26.0 for the independent, moderating and dependent variables, and the results of the hypothesis testing are shown in Table 7. The path coefficients are standardized estimates.

Table 7.

Hypothesis test results

Hypothesize	Estimate	P	Whether or not to support assumptions
H1	0.128	***	Support
H2	0.101	0.124	Unheld
H3	0.821	***	Support
H4	0.294	***	Support
H5	-0.971	0.868	Unheld
H6	0.528	***	Support
H7	0.528	0.676	Unheld
H8	0.737	***	Support
H9	0.877	***	Support
H10	-1.451	0.778	Uunheld
H11	-0.691	0.891	Unheld

Note: denotes P < 0.001

The standardized regression coefficient of perceived pleasantness on students’ intercultural communication competence after accepting the “intelligent learning system” is 0.101, with a P-value of 0.124, which is greater than 0.05. The standardized regression coefficient of “perceived pleasantness” on students’ intercultural communication competence after receiving the “intelligent learning system” is -0.971, and the P-value is 0.868, which is greater than 0.05. The standard regression coefficient of the intelligent learning system on students’ intercultural communication ability after receiving the “intelligent learning system” is -0.528, and the P-value is 0.676, which is greater than 0.05. The standard regression coefficient of the intelligent learning system on perceived ease of use is -1.451, and the P-value is 0.778, which is greater than 0.05. The standard regression coefficient of the intelligent learning system on satisfaction with learning is -0.691, and the P-value is 0.891, which is greater than 0.05. i.e., the P-values of the above five paths are all greater than 0.05, and the hypothesis test is not accepted, and the P-values of the remaining six paths are all less than 0.05, and the hypothesis test is accepted.

4

Conclusion

In this paper, we construct an intelligent language learning system based on the language learning data processing method of crawler technology and the intelligent voice quiz algorithm of similarity calculation model, and propose a research model of technology acceptance and use integration and 11 research hypotheses. Using structural equation modeling, a validation factor analysis was performed on the measurement model, and the structural model was tested for fit and corrected, resulting in a revised model. Eleven hypotheses were also validated, five of which failed the hypotheses and the remaining six passed the hypotheses. Perceived usefulness, community influence, facilitating conditions, and learning satisfaction positively affect students’ intercultural communication competence after receiving an intelligent learning system. Intelligent learning system has a positive effect on perceived usefulness and perceived effectiveness.

In conclusion, enhancing perceived usefulness, community influence, facilitating conditions, and learning satisfaction can help develop students’ intercultural communication competence.

Funding:

This research was supported by the 2021 Jilin Province Vocational Education Research Project “Research on the Difficulties and Countermeasures of Implementing the ‘1+x’ Certificate System for English Majors in Applied Universities in Jilin Province - Taking Jilin Business and Technology College as an Example” (Project Number: 2021XHZ037).

Funded by 2024 Ministry of Education’s Industry-Academia Cooperation and Collaborative Education Project:”Construction of Cross-border E-commerce Modern Industry College”

Sprache:: Englisch

Zeitrahmen der Veröffentlichung:: 1 Hefte pro Jahr
Fachgebiete der Zeitschrift:: Biologie, Biologie, andere, Mathematik, Angewandte Mathematik, Mathematik, Allgemeines, Physik, Physik, andere

Zeitschrift RSS Feed

Research on the Path of Enhancing Intercultural Communication Competence of Students in Applied Colleges and Universities Supported by Intelligent Learning System

Rui Shi

Online veröffentlicht: 17. März 2025

Eingereicht: 28. Okt. 2024

Akzeptiert: 12. Feb. 2025

DOI: https://doi.org/10.2478/amns-2025-0314

SchlüsselwörterTam model, utaut model, Intelligent learning system, Structural equation modeling, Intercultural communication competence

© 2025 Rui Shi, published by Sciendo

This work is licensed under the Creative Commons Attribution 4.0 International License.

Schlüsselwörter
Tam model, utaut model, Intelligent learning system, Structural equation modeling, Intercultural communication competence